ViTBackbone class

Inheritance

Constructors

ViTBackbone({required int imageSize, required int patchSize, int numChannels = 3, required int embedSize, int numLayers = 4, int numHeads = 4})

Properties

clsToken ↔ Tensor
getter/setter pair
embedSize → int
final
hashCode → int
The hash code for this object.
no setterinherited
imageSize → int
final
numChannels → int
final
patchProjection ↔ Layer
getter/setter pair
patchSize → int
final
posEmbeddings ↔ Tensor
getter/setter pair
runtimeType → Type
A representation of the runtime type of the object.
no setterinherited
transformerEncoder → TransformerEncoder
final

Methods

forward(Tensor patchifiedImage, List<Tensor> tracker) → Tensor
Takes a Tensor of shape numPatches, patch_pixels For a 224x224 image with 16x16 patches, input is 196, 768
noSuchMethod(Invocation invocation) → dynamic
Invoked when a nonexistent method or property is accessed.
inherited
parameters() → List<Tensor>
override
step(double lr) → void
inherited
toString() → String
A string representation of this object.
inherited
zeroGrad() → void
inherited

Operators

operator ==(Object other) → bool
The equality operator.
inherited